WPIA: accelerating DNN warm-up in Web browsers by precompiling WebGL programs

5 Conclusion In this paper, we study the long warm-up time of GPU acceleration of DNN inference in Web browsers. We analyzed the reason behind the long warm-up time through a measurement study and revealed that compiling WebGL programs takes most of the warm-up time. Inspired by this finding, we pro...

Full description

Saved in:
Bibliographic Details
Published in:Frontiers of Computer Science 2024-12, Vol.18 (6), p.186211, Article 186211
Main Authors: Tian, Deyu, Ma, Yun, Han, Yudong, Yang, Qi, Yang, Haochen, Huang, Gang
Format: Article
Language:English
Subjects:
Computer Science
Letter
Time measurement
Citations: Items that this one cites
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:5 Conclusion In this paper, we study the long warm-up time of GPU acceleration of DNN inference in Web browsers. We analyzed the reason behind the long warm-up time through a measurement study and revealed that compiling WebGL programs takes most of the warm-up time. Inspired by this finding, we proposed WPIA, an approach that suggests precompiling WebGL programs on the server side to avoid compiling them in Web browsers. WPIA tackles the challenges of precompiling by merging WebGL programs and using a record-and-replay technique. Evaluation experiment results show that WPIA can accelerate the DNN warm-up time to an order of magnitude.
ISSN:2095-2228
2095-2236
DOI:10.1007/s11704-024-40066-w